Product Engineer (Agents) (up to £120k)

Dex; Greater London; On-site; GBP 85,000 - 120,000; Full time

Job summary

Dex seeks a second Product Engineer for a new agents pod. You will own the Talent Agent Engine, shipping new agents and refining live ones, while building a production-grade AI stack for high-stakes hiring beyond a standard chatbot.

You will also develop the Studio control plane for agent configuration, experimentation, and feedback loops, and contribute to the voice pipeline for an Interview Agent in active build with live Scoring and Candidate Agents.

Qualifications

• Shipped AI agents in production and understand failure modes at scale.
• Deep experience with the modern agentic stack and stateful primitives.
• Designed memory, retries, instrumentation, and evaluation pipelines for agents.
• Full-stack range across candidate-facing UI, edge functions, and data storage.
• Curiosity for prototypes, benchmarks, and rapid learning over pure research.

Responsibilities

• Own the Talent Agent Engine: orchestration for intent, context, routing, tool use, validation.
• Ship new agents from roadmap to production and improve existing ones.
• Build the Studio control plane for agent configuration, experimentation, and feedback loops.
• Shape voice pipeline for Interview Agent alongside Scoring and Candidate Agents.
• Engineer for high-stakes workflows with focus on reliability and measurable outcomes.

Skills

AI agents in production

Agentic stack

Memory & retries

Observability & evaluation

Full-stack ranging

Tools

Orchestration frameworks

Job description

This role is with one of Dex's trusted partner companies. We work closely with their teams to truly understand their culture, goals, and what they're looking for, so we can match you with the right opportunity and give you context about the role before you commit to a process.

Dex is an AI recruiter agent that helps you run your job search. Tell Dex your stack, seniority, and what you want to build. We will manage your applications and surface other opportunities that are a fit.

The role

This platform sits uniquely between candidates and hiring systems, giving them a full view of the talent funnel. They're building a family of AI-powered Talent Agents to automate sourcing, scoring, scheduling, and interviewing. Backed by leading investors, they have significant traction, powering millions of candidate interactions for major enterprises.

This is the second Product Engineer for a new agents pod. You'll own the core Talent Agent Engine, shipping new agents and improving live ones. You'll also build the control plane for agent configuration and experimentation. This isn't a greenfield POC; you'll be building production-grade AI for high-stakes hiring, far beyond a typical customer-service chatbot.

The work

• Own the Talent Agent Engine: the orchestration layer for intent classification, context loading, routing, tool use, and validation.
• Ship new agents from roadmap to production, and continuously improve existing ones.
• Build the "Studio" control plane for agent configuration, experimentation, and feedback loops.
• Shape the voice pipeline for an Interview Agent currently in active build, alongside live Scoring and Candidate Agents.
• Engineer for high-stakes workflows where "looks right" isn't enough; focus on reliability and measurable outcomes.

What You Bring

• You have shipped AI agents in production, not just POCs, and understand failure modes at scale.
• You're deep on the modern agentic stack, having worked with stateful primitives and orchestration frameworks.
• You've designed memory, handled retries, instrumented observability, and built evaluation pipelines for agents.
• You have full-stack range: comfortable across candidate-facing UI, edge functions, and data storage.
• You prioritize shipping and learning; your curiosity leads to prototypes and benchmarks, not just research.